To the Editor.
—We would like to make 3 points in response to Dr Eddy's1 well-constructed article on what constitutes an investigational treatment. First, contrary to his conclusion, health plans should ocver investigational treatments, but only in an adequate learning context, that is, a scientifically adequate clinical trial or other research study or a quality management program with data collection mechanisms that can meaningfully illuminate patient outcomes in relation to care processes. The reason is simple: without such coverage we will never be able to learn which treatments are effective for whom.2 Second, while Eddy's technology assessment process may be ideal, there are too few pertinent studies and too few studies that produce usable results for it to work.3,4 Health plans must act in the present while we encourage the research establishment to increase the quantity and quality of its efforts
Goldschmidt PG, Monaco GP. Investigational Treatments: Process, Payment, and Priorities. JAMA. 1997;278(17):1402-1403. doi:10.1001/jama.1997.03550170032020